When a great genius appears, dunces can be relied upon also.

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Robert Flesher
Posts: 1280
Joined: Tue Aug 18, 2009 3:06 am

Re: When a great genius appears, dunces can be relied upon a

Post by Robert Flesher »

Romy wrote:In advance, forgive me my failures in English for my first post and all permitted successors.

I was attracted here by the controversy of early versions of RYBKA and any alleged small idea or even code/method borrowings or inspirations by its author Mr Rajlich.

You may not know that this matter even has escaped into the realworld media. Der Spiegel!!

I studied the whole thing, read a PDF from Mr Wegner and a PDF from another who did not directly sign his name to it, but who is known. And I read hundreds of post here and in open chess and hiarcs, and a hit-letter to ICGA with list of confederated signers.

I think by this standard applied against Mr Rajlich, even Mr Shakespeare should be banned. Shakespeare borrowed plots, methods, ideas even big groups of words. Plays are not chessprograms but the principle is similar.

Point is, Shakespeare's improvements were of great supergenius, not (relatively) mediocre like his so-called inspiration sources

More I studied this matter, more I was reminded of this wise quotation --

"This evil fortune, which generally attends extraordinary men in the management of great affairs, has been imputed to divers causes, that need not be here set down, when so obvious a one occurs, if what a certain writer observes be true, that when a great genius appears in the world the dunces are all in confederacy against him."
- Jonathan Swift, "Essay on the Fates of Clergymen", 1728

I do not call all the signatories to the anti-Rajlich as "dunces" but I do classify them as dunces by relativity - means by comparison to Mr Rajlich. In my opinion. And in opinion of Chess. Everything is relative. In their own circles I am sure they are all great wiseguys. Maybe, 20 or 30 years ago they too were genius not (relative) dunce.

Reason for my thinking is this. Their creations were not only beaten thoroughly by RYBKA, but were beaten with a maximum of humiliation! They could not even understand the chess method, how they were beaten! Most of the time they and their vast-inferior creations thought they were drawing, even winning, then as if my magic RYBKA prevailed. No magic. Just a higher chess.

So the 10 signatures to the anti-Rajlich letter, to me are all a nonsense and a confederacy. Of no appreciable value. No probative value. No evidential value. No big technical value. Only grapes value. Sour, even?

So what if Mr Rajlich looked at Mr Letouzey's work? He produce something of much higher magnitudes, level, understanding - Mr Rajlich's genius.

In all so-called evidence I see patch of code here and of source there. Nowhere I see understanding of how the alleged son is so much stronger (at Chess! not programming!) than the father.

If Mr Houdart (author of the HOUDINI) were to sign such a hit-letter as produced against Mr Rajlich, then I would pay some attention to the letter. At least HOUDINI is a suitable calibre opponent to RYBKA, its author's opinion deserve respect. All I can see is result of RYBKA-CRAFTY or other lacklustres. How many game I have to play from neutral openingbook before Crafty win single game against RYBKA? 300? 3000? More? JUNIOR same.

I am waiting to be sent testcopy of Dr Wael Deeb's chessprogram. His most gentle, most refined language is a good attracter for me.

History will record this attack event much the same as my analysis, of this I am confident.

ICGA better disband before Mr Levy gets into trouble, I would advise. It has no legal role and should watch carefully unwise statement.

My advice to Mr Rajlich, ignore the mob, I mean confederacy. Let them make sanctions. History is to your side. Tournament without RYBKA is like human chess world without KASPAROV. Do not even soil your hand with lawsuit. It is in the inherent(?) nature of dog to bark when it sees a frightening power, and of dunces to confederate, so there is no question of blame.

Even if Mr Rajlich did 90% of what he is accused, no big deal, it is matter of legal interpretation. Mozart had but once to hear other's music, he internally recorded and processed then improved greatly. There are many examples.

Condemn Rajlich, you condemn Shakespeare and Mozart too.

Excuse me, I have no intention of providing offense to any person, even Salieri. Only for perspective. So, confederacy of (relative) dunces, please excuse my statements, they all were given with the maximum of due, earned respect to you.

Romy

Carol as it now seems clear that you are a "pro-angler", I thought I'd share this passage with you.


" As the calendar turns to May, I always think of the 24th of May fishing trips planned by my father and brothers.

Almost every year, Dad would take the three oldest boys and head out for "ponds unknown" to do some troutin'. Off they would trek while the skies were still dark. They would often head out over the Trans Canada Highway and pull off near Soldier's Pond. They'd don their long rubbers and walk about 5 miles, over muskeg and shrub brush, to get to a pond "that was not over-fished, you know". I hear my brothers reminiscing now and the tales are all of rooting each other in the behind with the tips of their fishing rods as they trudged along over the barrens and skylarking until Dad would get mad at them. The oldest boy would complain as he would have to lag at the back to make sure the younger ones didn't get too far back and get left behind.

The night before the trip Mom would make a stack of potted meat sandwiches (with a bit of mustard pickles diced up in it for flavour), wrap them up in waxed paper and thrust them into an old canvas back pack Dad had. Of course there was a small kettle and box of Seadog matches tucked in along side for the boil up along the way. A tin of beans set over the open fire to heat and the little kettle on to boil having been filled with water from the nearest brook… I don't know if there is a cup of tea in the world that tastes as good as one made over an open fire with blasty boughs and juniper twigs firing a kettle full of brook water!

Another popular 24th of May trek for our family was out to a place near the Witless Bay Line. There was a family, named Kane, who used to be our neighbours in the early days. Old Mr. Kane had a summer shack out Witless Bay area that had a renowned spring giving them drinking water. The spring must have come from deep in the Earth as the water was as cold and pure as any our can find anywhere. The Kane family was devout Catholic and their legend was that a priest, who was a relative, had blessed the well when it was first dug and it flowed pure water due to this. Dad knew of a good fishing gully in the general direction of the Kane property so they would stop at the spring for a "mug-up" and then trek another 3-5 miles to the place he wanted to fish.

Once they got where they were going (Dad always seemed to have a place in mind back then, but now the boys think it was all a ruse) Dad would get them set up and he would go to another side of the pond to wet his line. Of course, as soon as the afternoon sun got at them; after the early rise and long walk over the barrens, they would often as not lie down on the moss for a snooze. Mr. Kane joined them for some fishing one year and Bill recalls how much fun it was to watch him. The old fellow was convinced that the fish could see you up through the water so to get the big ones to bite, you had to sneak up on them to cast out your line…so as not to scare them out into deeper water. He said the old guy would get into a queer crouch about 4 or 5 feet from the side of the pond. He would then scuttle forward, still in his crouched position and fling his line out. He would stay crouched until the line got slowly reeled back in and then repeat the procedure. Bill and the other boys laughed and thought the man quite mad until he started reeling in the fish. He got the best catches of the day. That gave them something to think about as they trudged back across the barrens!

One year, my Father took the Newfie Bullet "Trouter's Special" with my oldest brother, Bill and his best friend, Eric. Dad reckoned there was no where else on Earth that a man could take a free train ride to some of the best wilderness fishing and back, but that's what you could do on the Trouter's Special on May 24th weekend.

For those of you that don't know, the Newfie Bullet was the name of the trans-provincial train that used a narrow gauge rail and would plod its way slowly back and forth across the Island from the late 1800's until the 1980's when it was deemed to be useless and the tracks torn up…to be replaced by a trans-provincial bus service run by C.N.

The Newfie Bullet moved at such a slow pace that the standing joke was about a woman who went to the conductor when the train neared Corner Brook and told him he had to stop the train as she was about to deliver her baby. When he chided her for attempting such a trip in her condition, she informed him that she was not pregnant when she left St. John's! Anyway, being a British colony until 1949, Newfoundlanders took their 24th of May holiday (celebration of Queen Victoria's birthday) seriously. The Bullet made a special run that weekend from St. John's out across the Avalon Peninsula as far as Argentia, to take folks to good fishing spots. You could pretty well tell the conductor to stop anywhere along the line by pulling the rope bell. The Trouter's Special left St. John's on Friday afternoon, with a lot the men weighed down by their fishing baskets full of Screech and beer, and it returned Sunday afternoon. The trip out was apparently quite a lot of fun with the men singing and telling tales of May long weekends past (the fish were HUGE, don't you know!) For passage back to St. John's you had to be sure you were by the side of the tracks to flag the train down as it went past on Sunday. You had better not miss the return trip, though, or you'd have a long walk ahead of you!

I have been known to wet a line occasionally over the years but I didn't get the "fishin' bug" like some of my siblings did. One of my daughters is an avid angler, though. She often spends at least one afternoon fishing on the "May long".

However you spend your May 24th weekend, do something you really enjoy doing with people you like spending time with! Make it a celebration of spring, be safe and most of all….be happy! "


Now we have both shared great "fish" stories, I bid you good luck in your future *angling* adventures.
Terry McCracken
Posts: 16465
Joined: Wed Aug 01, 2007 4:16 am
Location: Canada

Re: When a great genius appears, dunces can be relied upon a

Post by Terry McCracken »

geots wrote:
Romy wrote:
geots wrote:Welcome to the forum. It may be too late, but I hope you had all your shots and other vaccinations before posting here. There is a lot of crap floating around.
Thank you. I did. And I observed as you said.

Even, it appears I am some person or persons I am not. Same people who made the false accusation against Mr Rajlich have made baseless, evidenceless accusations against me. Making accusations without even the semblances of proof is apparently a norm here, and some moderators either participate or permit.

Mr Rajlich always appeared a gentleman in posting. But he has invoked a massive inferiority complex in some other. Chess is only the game.

Since two apparently very different source codes can, given supercompilers, render into the same object code, comparing object code of RYBKA (early) and other programs (source or object, it does not concern) proves...nothing, except about the logic powers of the pack hunters.

Showing that Fruit source in small parts compiles into object code modules similar or even identical to chunks within RYBKA is an example of non-proof. S/N ratio high to overload.

Watch your back. People are beginning to realize you have more knowledge in programming with each of your posts. When they cant attack your premises- the Vas haters will attack your character. Isn't that the way a democracy works?
He knows didlysquat and you know less and welcome the Snake into the Garden.

Get a life people!
Terry McCracken
User avatar
Romy
Posts: 72
Joined: Thu Mar 10, 2011 10:39 pm
Location: Bucharest (Romania)

Re: When a great genius appears, dunces can be relied upon a

Post by Romy »

geots wrote:People are beginning to realize you have more knowledge in programming with each of your posts. When they cant attack your premises
Mine are not premises! They are facts, and even demonstrable in any competent computer laboratory.

The source code of RYBKA, and not what they think might have been the source code of RYBKA on the basis of decompilation adventures, is needed to establish their claims even 1%. Otherwise it is conjecture based on their incorrect premise, easily disproved by experiment, that decompilation guarantees to source congruency. It does not. Many DIFFERENT sources can be put through the SAME compiler and produce IDENTICAL object code.

They do not have this source code. Due to accident - so why they waited five years - they cannot ever have it, as it no longer exists except partly in the human memory of Mr Rajlich due to some backup mishap (only my memory, I read it somewhere).

Do not lose any sleeps over this nonsense. First time proper computer scientist gets his teeths into the question, baboon pack will be scatter.
Milos
Posts: 4190
Joined: Wed Nov 25, 2009 1:47 am

Re: When a great genius appears, dunces can be relied upon a

Post by Milos »

Romy wrote:Mine are not premises! They are facts, and even demonstrable in any competent computer laboratory.

The source code of RYBKA, and not what they think might have been the source code of RYBKA on the basis of decompilation adventures, is needed to establish their claims even 1%. Otherwise it is conjecture based on their incorrect premise, easily disproved by experiment, that decompilation guarantees to source congruency. It does not. Many DIFFERENT sources can be put through the SAME compiler and produce IDENTICAL object code.
You can give a keyboard to many different types of monkeys, like chimps, gorillas, baboons... and all of them would produce exactly the same kind of random BS as you do...
If your BS is put through some compiler, it would still be the same type of BS, so your thesis is proven.
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: When a great genius appears, dunces can be relied upon a

Post by bob »

Romy wrote:
bob wrote:the "samplings" involved verbatim copying of (source) code
And your copy of the RYBKA source to demonstrate this is...where? Is this an example of prosecutorial standards? Or lynch mob?
A computer scientist does not need source. Binary will do just as well. And it is crystal clear in revealing the massive copying from Crafty in the pre-1.0 beta versions (1.x thru 1.6.1)...


Let me give you the big hint.

Compiler and Assembler are very different tools.

Assembler is a trivial. Just 1 on 1 conversion, few assists, nothing special. If you have object, you have source (minus some labels). If you have source, you have object. Machine language with mnemonicals.

Compiler, in contrasts, is non-trivial. Translates from human high to machine low. And therefore, no even close 1 to 1 conversion. Instead, many to 1. This means, depending on intelligence of compiler modules, many different sources can compile to exactly the same object (code). This is because of the OPTIMISE power of the compiler.
I can know this was not the case so much in 1616 but Rybka is from 2006.


And you don't think that someone with experience can take a highly optimized binary, and compare it to a C source, to determine if they are semantically equivalent. For years, I had FORTRAN copies of many functions in Cray Blitz that had been completely rewritten into incredibly well optimized assembly language. When we made a change, we first changed the FORTRAN, and used that rather than the asm to check the effect. If we were happy, we then modified the assembly. And it was _always_ possible to do this, even though the two did not look anywhere near similar.

Just because YOU can't do it, doesn't mean that nobody ELSE can.



Even to make an absurd extension (no application here) for those with comprehension disability, an advanced compiler, could have a library function eval_fruit, so the author just calls the function and the object is produced automatically.


So? One can look at the library binary just as easily. Although we have not considered library code as relevant so far since there is no chess-specific stuff in them, today.


It is easy to show that advanced compilers, and depending on their libraries and settings, can produce identical object code from widely-differentiated source code. Optimisation include automatic rearrangement, reorder, reprioritise, replacement, removal.


Have you ever _written_ a compiler? I have written several. I have worked on others. I have written an assembler. Three, in fact. One for the IBM 360, one for the old Zilog Z80, and one for an HP 2100a microprogrammable mini.

Re-ordering, instruction lifting, common sub-expression elimination, strength reduction, and such are all well-known techniques., They change _nothing_ with respect to semantics. I have no idea where you are going with this, but it is not going anywhere helpful to the case of Vas.



Some table values same is all this is reduce to.

Conclusion -- no source, no conclusion, no prosecution or witch-hunt Alabama lynching. Just allegation, noise and potential for trouble and backfire.

As Prof.(ass.), I can imagine you already know at least basic the operation of compiler. Please then you should educate other confused signatories.
Apparently I know a whole lot more about compiling, assembly language programming, and assembly language optimization than you do. Based on the totally off-base comments you are making. There are two issues. One is _identical_ code sematically. The other is identical strings for output. Do you believe that Vas and I magically just chose the exact same _identical_ output formats for things like the benchmark command, the perft command, and such? That we picked the same 6 positions to search for the benchmark command. There are so many levels where this is wrong there is nothing to discuss...
bob wrote:The only "dunces" here are those offering such inane and ridiculous arguments to justify that which simply can not be justified
Dunces or Baboon, you can choose the better cap. I have less qualification.
bob wrote:We are talking about copying, _verbatim_ large pieces of both Crafty _and_ Fruit. Shakespeare did _not_ do that.
I understand we are not in Alabama. Here are some dates--

Crafty born after 1986, as son of Cray Blitz. Fruit born even later, borrowing from many places.
Get it right. Crafty started in late 1994 after the last ACM CCC event. It played its first game on ICC in either late December 1994 or early Jan 1995, not sure which. Prior to that it was all "Cray Blitz"


But Shakespeare died few years before, in 1616.

It is safe to make a conclude that Shakespeare did not copy verbatim large pieces of both Crafty and Fruit.
Got a point there, although I don't see how it helps Vas. Just Shakespeare.
bob wrote:And there is some sort of point to the post and picture? Eludes me.
Is it necessary to explain to a Professor (ass.) that vandalism is not to be encouraged? Even, of reputations?
Don't see any valdalism, just "exposure of the truth" which is something entirely different...
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: When a great genius appears, dunces can be relied upon a

Post by bob »

Romy wrote:
Robert Flesher wrote:we have respected "experts" in field of computer chess stating there is illegal code copying present in Rybka 1.0.
So-called "experts" in field of computer chess may not be such experts in field of optimising compilers...hopefully, they can extricates themselves before it bites them. See my contribution above for relevance.
There is _absolutely_ nothing that has been said that will bite anyone but the person that copied the code in the first place. To believe otherwise would be stupid. Since I don't believe you are stupid, you are therefore more than a bit dishonest. No other explanation for such ridiculous statements.

I am easy to find. Sue me.
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: When a great genius appears, dunces can be relied upon a

Post by bob »

Romy wrote:
geots wrote:Let's see the code for Shredder and Hiarcs. I would like to see how they got within 200 elo of Rybka.
Hear, hear!

Stone thrower pack is apparently less aware that powerful weapon can be deployed against them. Psychopathology of baboon is brave in collections, until a Abrams M1A2 roll into view. Then I predict a scattering.

I checked my compiler reasoning with a computer department of top university. And object can be a patch of output chunks from different compilers with I/O defined. I am so informed.
Simple response. "Bullshit".

You might better ask them again and mention the expression "semantic equivalence between a C source and an optimized ASM source. " If they can't address that, they are not computer scientists... not even close.
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: When a great genius appears, dunces can be relied upon a

Post by bob »

Romy wrote:
geots wrote:you could I suppose be wasting your time. Not much you can say to someone like Bob who agrees to some sort of screwy hearing over Vas and Rybka, then states that "even if he is found innocent, he is still guilty". Of course they all know that they cant prove anything unless they get their hands on the code for any versions of Rybka AFTER Beta 1
They do not have even bonafide source for RYBKA Beta 1 (only Strelkafied reverse engineer job, so it fall foul of non-reversibility caused by many-to-one relationship between high-level source and object), so my refutation still applies.

If this gets within ten kilometres of the correctly constituted judicial authority, the many-one nature of compile process will deliver terminal blow to stonethrower logic. To win Mr Rajlich has to prove nothing, a compile process from undefined compilers' modules is irreversible functionally, period. The stonethrower must show other-wise. This is impossible.

All that "decompile" can produce is a source that may have been legally congruent to the original, and which when compiled will produce the object code. In no way it suggests, let alones proves, that this was legally congruent to the actual source used by Mr Rajlich.

From the university of Mr Rajlich's family can come experts to testify this, experts real at whose name the ICGA will tremble. Demonstrations can be arranged with actual compilers of very different source compiling to identical object.

Then Mr Rajlich can choose who has most money from any vandal pack, and obtain a compensation by due process.

But, it will not become to this. Mr Rajlich is too gentleman and Mr Levy too clever to fall inside the baboontrap. (I have no authority from or connection of Mr Rajlich, last sentence is my hope for community good. I derives my authority only from legal practice).
Again. "total bullshit". We are not taking assembly and working back to C to prove that the two are identical. We are analyzing the C of Crafty (or fruit) and comparing to the assembly of Rybka, and proving semantic equivalence. That is ironclad. It has been used in many universities around the world to detect disguised plagiarism. Feel free to try to sue anyone you want. It will get real expensive, real quick. As you open the door for reverse suits at the same time, and those will be painful.
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: When a great genius appears, dunces can be relied upon a

Post by bob »

geots wrote:
Romy wrote:
geots wrote:Welcome to the forum. It may be too late, but I hope you had all your shots and other vaccinations before posting here. There is a lot of crap floating around.
Thank you. I did. And I observed as you said.

Even, it appears I am some person or persons I am not. Same people who made the false accusation against Mr Rajlich have made baseless, evidenceless accusations against me. Making accusations without even the semblances of proof is apparently a norm here, and some moderators either participate or permit.

Mr Rajlich always appeared a gentleman in posting. But he has invoked a massive inferiority complex in some other. Chess is only the game.

Since two apparently very different source codes can, given supercompilers, render into the same object code, comparing object code of RYBKA (early) and other programs (source or object, it does not concern) proves...nothing, except about the logic powers of the pack hunters.

Showing that Fruit source in small parts compiles into object code modules similar or even identical to chunks within RYBKA is an example of non-proof. S/N ratio high to overload.

Watch your back. People are beginning to realize you have more knowledge in programming with each of your posts. When they cant attack your premises- the Vas haters will attack your character. Isn't that the way a democracy works?
Actually, isn't that the way you and Chris W have worked??? You have it backward...

The evidence is ironclad, so you now attack the people gathering the evidence instead.
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: When a great genius appears, dunces can be relied upon a

Post by bob »

Romy wrote:
geots wrote:People are beginning to realize you have more knowledge in programming with each of your posts. When they cant attack your premises
Mine are not premises! They are facts, and even demonstrable in any competent computer laboratory.

The source code of RYBKA, and not what they think might have been the source code of RYBKA on the basis of decompilation adventures, is needed to establish their claims even 1%. Otherwise it is conjecture based on their incorrect premise, easily disproved by experiment, that decompilation guarantees to source congruency. It does not. Many DIFFERENT sources can be put through the SAME compiler and produce IDENTICAL object code.

They do not have this source code. Due to accident - so why they waited five years - they cannot ever have it, as it no longer exists except partly in the human memory of Mr Rajlich due to some backup mishap (only my memory, I read it somewhere).

Do not lose any sleeps over this nonsense. First time proper computer scientist gets his teeths into the question, baboon pack will be scatter.
Please stop posting that bullshit. A piece of code expresses a semantic idea. Yes, there are many different ways to express the same semantic idea, such as.

For (i=0; i<100;i++) {
...
}

or

i=0;
while (i < 100) {
...
i++;
}|

or

i=0;
loop:
if (i >= 100) goto exit;
...
i++
goto loop;

or

<etc>

But they are all semantically _identical_ and produce the same asm when optimized to extreme levels.

What is the probability that two different programmers will produce the same C source? Near zero. What is the probability that two different programmers will produce different source that do the same thing semantically? Even lower. That is what you don't get. Or, actually, you probably do get it, but want to try to sound like you know what you are talking about while actually blathering like an idiot. And it gets worse when you copy someone's source, which has a known bug in it, and then that same bug is found in the assembly language produced by the compiler. Pretty tough to explain that one away with the blathering garbage you are providing. But keep trying.

It does keep the internet bandwidth suppliers busy.